WSP00036

• File Number: 8200.750
• Description: San Juan River General
• State: CO
• Basin: San Juan/Dolores
• Water Division: 7
• Date: 3/1/1975
• Author: BOR
• Title: San Juan Ecology Project - Interim Progress Report - March 1975
• Water Supply Pro - Doc Type: Report/Study

I

001583
a variety of the measured environmental data were
submitted to factor analysis which delineated two
major patterns of variability, which will be referred
to as Factor 1 and Factor 2. Factor ,1 represented
all those variables which varied diurnally with solar
radiation input and were associated with atmospheric
evaporative demand and energy budget of the tree
crown. Factor 2 represented the deep soil regime
which varies on a seasonal basis and included soil
water and deep solI temperature. Water stress was
primarily associated with Factor 1, but was affected
to some degree by Factor 2. The following conceptual
linear model is proposed:

Water Stress ~ Bo + Bl (atmospheric factor ) and B2
(deep soil factor) + E1j.

Correlation analysis showed high individual correla-
tions between water stress and all Factor 1 variables,
with individual R2 values up to .68, but extremely
low correlations for Factor 2 variables. The four
best individual variables were ranked as follows:
surface soil temperature:> air temperature :>vapor
pressure deficit~solar radiation. This ranking may
have indicated greater dependence on sensible energy
and leaf temperature than on evaporative demand. The
two, two-variable combinations with highest correla-
tions included both a Factor 1 and a Factor 2 varia-
ble. It is important to note that Factor 2 greatly
increased in importance in the presence of Factor 1.
This substantiated the proposed model.

I

I

I

I

I

I

I

I

Prediction models were then constructed by stepwise
multiple regression based on the conceptual model.
Significant models with R2 values up to .87 were
produced with up to nine environmental variables.
A practical four-variable model with two variables
representing each factor was finally developed with
an R2 = .81. The four variables were surface soil
temperature, and -15 cm soil water content. This
model is ecologically sound because water loss and
water absorption are both represented, water stress
being the sum of the difference between the two
processes. The model can be used for predicting water
stress through the midsummer season from basic weather
data, as long as it is not extrapolated beyond the
range of the original data.

I

I

I

Snowmelt Period Water Stress

I

I

I

I

I

'\,.....

I

'OH" '"I" ouG< ""
Figure 1. Seasonal tree mOlsture s.rrese, ma.x..llllUUI
and minimum as related to snow depth and cover.

I

By examination of the figure, the relationship to
snow becomes apparent. Minimum (predawn) water
stress remained fairly constant across the snowmelt
period which is probably due to a fairly constant
high level of soil water. This demonstrates that
the trees were always able to recover satisfactorily
from midday water stress during the period. The
most significant information of Figure 1 is the
maximum daily water stress. It is apparent from the
figure that maximum daily water stress decreased
almost 50% over the snowmelt period in a fairly
linear manner. It is also important to note that
this decrease is highly correlated with the decrease
in snow cover (R2 = .62) and snow depth (R2 = .74).
It is probable that this phenomenon is an expression
of the second hypothesis presented above, i.e., cold
soil temperatures increase the root resistance to
water flow which causes higher water stress levels.
As the snow melts and more soil becomes exposed to
radiation, the average soil temperature across the
site increases, allowing root resistance to decrease,
water uptake rate to increase, and midday water
stress to decrease.

As previously stated, water stress is an expression
of the difference between water loss through transpir-
ation and water uptake through root absorption.
During the snow melt period, transpiration should be
increasing due to the increasing evaporative demand
of the atmosphere (not shown in Figure 1). Since
water stress decreases, it must be assumed that
uptake is not constant, but is actually increasing
more rapidly than transpiration. As any change in
soil water would cause a water stress change in the
opposite direction, it might be concluded that the
pattern exhibited during snowmelt is possibly due to
some intrinsic factor, however, it is more likely a
result of increasing soil temperatures as more of the
soil surface becomes free of snow.

The root resistance laboratory study, using Engel-
mann spruce seed under simulated snowmelt
conditions was initiated to determine the validity
of this assumption. Preliminary examination of the
results indicates thab the mean rate of water move-
ment into and through the root systems was approxi-
mately three times greater with water at 7.5oC as
with water at a.soC. The mean rate at lsoC was
slightly higher than the 7.50 rate, but the variance
was exceptionally high. These results substantiate
the conclusion that the high water stress levels
under snowpack are caused by soil temperatures near
freezing during periods of high transpirational
demand.

The implications of these spring water stress patterns
to tree growth are very important. Since the trees
are able to recover from water stress at night, no
immediate severe effect on cell division and cell
enlargement would be expected due to delayed snowmelt.
The primary effect would likely be a significant
reduction in photosynthesis. Many investigators
have shown that photosynthesis is reduced to zero at
water stress levels in the range of -15 to -25 bars.
This could have a significant effect on growth later
in the current season and on the subsequent season's
growth. The timing of the major phenologic events
will be very important in determining if this high
stress period will have an important effect on growth.
If cambial initiation and bud burst are related to
environmental temperatures and/or water stress, they
will be delayed due to increased snow, although trees
may be able to adjust their growing season. However,
this does not appear to be the case. If these events
are controlled by photoperiod and occur at approxi-

26